Looking at the lists with that in mind, you're right. It's worth discussing whether you want that though. A swarm deck has to comprise of a lot of cheap minions and I'm not sure there's any way to get that together here. I count around 35 cards that'll give you a minion costing 2 or less (including cards which pull other cards and minions that can play for less in some circumstances like Goblin Scavengers with Goblin Armory), out of a pool of, what, 234 cards? Only 10 of those could cost 1 or less. Swarm doesn't usually need very much synergy at all, stupid swarm (playing 1-cost minions regardless of culture) has won tournaments. In this pool of cards, though, you probably do need a lot of it. With that in mind, Goblin Armory doesn't seem very impactful after all. Strange to say, but true.
When I said swarm requires synergy, I meant that swarm cards (like 1-cost minions) aren't very powerful unless you play a lot of them. "Mostly swarm with a splash of beatdown" works, but "mostly beatdown with a splash of swarm" doesn't.
One thing I'm embarassed to say I hadn't considered was the spread of 1-cost minions over the whole shadow pool; I mostly looked at curves one culture at a time. I think it would make a lot of sense to add more 1-cost and 2-cost minions in other shadow cultures so that a Moria-based swarm deck has more options to splash from other cultures. That, combined with dedicating 1/4-1/3 of the deck as beatdown to slow the opposing fellowship, should make swarm strategies far more plausible.
All of that assumes that 7 minions on the table is competitive in a format where people are probably not too attached to any particular companion and have little to lose if Enquea kills one. There's a reason cards like Far Harad Mercenaries were printed as commons. If I didn't do anything wrong, swarm decks aren't really going to happen. Now, maybe that's good. I really disliked playing against swarm for my first few years in the game, and it means you don't need to worry nearly as much about RB support.
I don't think swarm is ever going to be the most reliable/powerful strategy in this format, but I would like it to exist. Using very rough numbers, I think having it be a strategy that works in 5%-10% of pools (in other words, around 1/2 or 2/3 of Moria pools) would be a good goal. That way players can't ignore RB support, but can still be competitive if they don't open much.
The power level and frequency of large fellowship hate is something absolutely needs to be fine-tuned by someone with more competitive experience than me.
Now, I agree with your hesitancy to just up the number of packs. It's not really sealed when you've got all the cards you want anyway. At the same time, without enough cards to build towards some direction it's still not really sealed. I think there's not enough room for strategy in the current layout. But again, I might be looking at this all wrong. You've done some simulations, maybe you can share a sample deck? Or perhaps PM me the code you used so I can run a few for myself?
I've tried to attached the code to this post; if that doesn't work, I can dm it to you. It's a short python script (CreatePool.py) that uses the other two attached files (Template.txt and MasterCardlist.csv) to create a file that can be imported as a format into Zorbec's Decklist Builder. The "format" is one in which the only legal cards are the ones that you opened (including limits based on how many you opened). I'll also pop a couple of examples as a follow-up post for people who are interested to see what they look like without having to create their own.
Another option for a "base pack" is to format it the same way as the others. A group of 36 cards where you get 11 of them based on rarity. That's what I was thinking, at least, although again it might require the other packs to be restructured and I don't think that's a good thing at this point. But you can ensure that these are characters with little to no cultural enforcement and that the rares are solid starting companions or somesuch. I have not looked at companions the same way I looked at e.g. swarm above though, it's quite possible that things are fine now. That's why I was saying prioritize splashability for Free Peoples commons though, if a player only has 50 reasonably-usable cards then deckbuilding is more about detail than direction.
I haven't found problems building a functioning FP side, but I would love to hear what you think after looking at some pools yourself!
Ultimately, one of the reasons Draft works so well in Magic is that you can make your deck work after you get your cards. You don't really draft the core, you draft that cards you want to go around that core and then choose the core yourself. The core for Magic, of course, is lands. The draft pack and Hobbit Draft Game base deck basically serve this purpose, just provide it up front. You know what you're building from in those cases, and in Magic where you know what you're building toward. Either way, there's an anchor for it all. I think FPs in particular really need an anchor.
This is a really interesting perspective! The difference, of course, is that lands are deliberately simple/uninteresting, whereas starting companions are hopefully exciting. A better analogy might be the commander in a Commander Legends draft: ideally you pick an exciting one during the draft, but there's always Prismatic Piper to fall back on if you have to. That leads me personally towards having some "basic" versions of companions available to everyone to make sure they can always fill out a deck. It's still something I'd prefer to avoid, but it deserves some serious thought.
This is a tangent, but there's also an as-of-yet-unused part of Gemp's draft where you pull multiple cards at once. Players could have a choice between, for example, Legolas with two weapons vs Aragorn with his two weapons vs Gandalf with his sword and staff. It would be awfully complicated to wedge such a thing into this concept, I just felt it was worth mentioning for some reason. I guess we've talked about every other way to spice up sealed already so it seemed to fit
As you say, this would be complicated to wedge into this concept, but it seems like a very efficient way to make narrow support cards like Ranger's Sword work in a draft environment.